Summary

The current forecast date is 28 March.

The forecasting model is an unweighted ensemble of three individual models:

  1. unweighted time series ensemble model (autoregressive ARIMA, ETS and naive models)
  2. regression + 7-day-lagged cases (cases mapped from UTLA to Trust using Trust-UTLA mapping in covid19.nhs.data)
  3. convolution from cases to admissions (scaled; cases mapped to Trusts as above)

All models are trained on the last 6 weeks of data (from 14 February) and forecasts are made for the following 14 days.

Current forecast

By recent observed admissions

By forecast admissions

Probability of increase (scatter)

Probability of increase (map)

Past forecasts